Social media has become a valuable source for market signals in the digital assets ecosystem, and companies like Solume are at the forefront of sentiment analysis. Leveraging their expertise in NLP and domain knowledge of cryptocurrencies and digital assets, the Solume team built a social media analytics platform to aid in predicting price movements. Solume caters to hedge funds and financial institutions and we are happy to announce that the Solume data set is now available on Blocktap.

Solume tracks social media sentiment across Reddit, Twitter, BitcoinTalk and other sources. They apply custom filtering and analysis to remove fake signals and bots. Each asset is assigned a sentiment score that changes over time based on social volume activity. For integration into Blocktap, we made historic Solume data available through time series queries. This allows users to query market price data and Solume sentiment data in a single query.

The following example is a time series query that fetches the price and sentiment data for Bitcoin and Litecoin in the past 24-hours. The time series resolution is set to 1-hour, which is the smallest time interval available in the Solume data set:

query price_sentiment {
  timeseries(resolution: _1h, limit: 24) {
    startUnix
    startDate
    
    # add BTC/USD and LTC/USD price data
    markets(filter: { 
        baseSymbol: { _in: ["BTC","LTC"] },
        quoteSymbol: { _eq: "USD" }
        exchangeSymbol: { _eq: "Bitstamp" }
      }) {
      marketSymbol
      closePrice
    }
    
    # add Solume sentiment data for BTC and LTC
    assets(filter: { assetSymbol: { _in: ["BTC","LTC"] }}) {
      assetSymbol
      sentiment {
        score
        socialVolume
        authenticSocialVolume
        positiveSocialVolume
      	negativeSocialVolume
        botsSocialVolume
      }
    }
  }
}

This query returns a time series array of JSON objects for each hour that includes both market data and Solume sentiment metrics:

{
  "data": {
    "timeseries": [
      {
        "startUnix": 1550264400,
        "startDate": "2019-02-15T21:00:00.000Z",
        "markets": [
          {
            "marketSymbol": "Bitstamp:BTC/USD",
            "closePrice": "3565.10000000"
          },
          {
            "marketSymbol": "Bitstamp:LTC/USD",
            "closePrice": "41.70000000"
          }
        ],
        "assets": [
          {
            "assetSymbol": "BTC",
            "sentiment": {
              "score": 0,
              "socialVolume": 157,
              "authenticSocialVolume": 41,
              "positiveSocialVolume": 0,
              "negativeSocialVolume": 0,
              "botsSocialVolume": 40
            }
          },
          {
            "assetSymbol": "LTC",
            "sentiment": {
              "score": 0,
              "socialVolume": 4,
              "authenticSocialVolume": 0,
              "positiveSocialVolume": 0,
              "negativeSocialVolume": 0,
              "botsSocialVolume": 3
            }
          }
        ]
      }
      
      ...

Click here to run this query in the Blocktap query tool and view the full result set.

Time series queries are an incredibly powerful feature in Blocktap and they are about to get even better as we continue to index additional data sets. Stay tuned!